Clinical and radiological outcome of surgery for pyogenic and tuberculous spondylitis: comparisons of surgical techniques and disease types

Abstract

Object Current surgical techniques for patients with pyogenic spondylitis (PS) and tuberculous spondylitis (TS) are as follows: anterior debridement/decompression and fusion with bone autografts (A); anterior debridement/decompression and fusion, followed by simultaneous or sequential posterior fusion with instrumentation (AP); posterior fusion with instrumentation, followed by simultaneous or sequential anterior debridement/decompression and fusion (PA); and posterior decompression and fusion with bone autografts (P). In the present study the authors compared, between disease types and between surgical techniques, the clinical and radiological outcomes of surgery for these patients. Methods Fifty-two patients were involved in the study, comprising 25 with PS and 27 with TS, with a mean age of 63.3 ± 13.3 years. The affected sites included cervical vertebrae in 6 patients, thoracic in 16, thoracolumbar in 14, and lumbar in 16. Surgical techniques A, AP, and PA were each performed in 15 patients (designated Groups A, AP, and PA), and technique P was performed in 7 patients (designated Group P). Clinical and radiological outcomes were evaluated between disease types and surgical techniques. Advancement on the Frankel scale between preoperative and follow-up scores was used as the criterion of neurological recovery. Results There was no difference in neurological recovery between disease types; however, the period of hospitalization was significantly longer in patients with TS than in those with PS. There was no difference in correction rate and loss of correction between disease types. Prolongation of the duration of disease was associated with a significant decrease in neurological recovery in TS, and the same tendency was observed in PS. It was also found that prolongation of the interval to negative C-reactive protein findings was associated with a significant deterioration in neurological recovery in TS. Also, no difference in neurological recovery was found between surgical techniques. Favorable degrees of correction were obtained in all groups after surgery, and favorable alignments were maintained until follow-up in both AP and PA groups, in which instrumentation was used. On the other hand, in Groups A and P, in which no instrumentation was used, correction losses of 4.5 ± 1.4°and 3.5 ± 2.7°, respectively, were detected at follow-up. The period of hospitalization was significantly shorter in Groups AP and PA compared with that in Groups A and P. Conclusions It was demonstrated that prolongation of the duration of disease or interval to negative C-reactive protein findings was associated with poor clinical outcomes, suggesting that surgical treatment should be performed without hesitation for patients unresponsive to conservative treatment, those with neurological symptoms, and those with kyphosis. Regarding surgical techniques, AP and PA can be recommended because they provide a significantly smaller loss of correction and a shorter period of hospitalization than those without instrumentation. There were no differences in clinical or radiological parameters between Groups AP and PA, indicating that either of these 2 surgical techniques may be selected flexibly depending on the patient's condition.